One of my advisors asked me a question that cut through a bad roadmap. Why are you helping dinosaurs run faster?
I was mid-pitch about going into legacy businesses to install AI and upgrade their operations. The dinosaurs have capital. They need to spend it to survive. They need what we know. On paper, it is a real market.
The question rearranged the roadmap anyway. Running faster is the wrong outcome for a dinosaur. The dinosaur needs to stop being a dinosaur, and no consultant can do that operation.
Three Lanes
There are three lanes in the AI economy right now. They look similar from a distance. They do not pay the same.
Lane 1: Dinosaur enablement. AI consulting, workflow automation, custom GPTs for legacy businesses. Incremental revenue. Most of the work is education rather than implementation: the sales cycle is "let me explain why you need this first." The leader you are selling to is reflexively scared of what you are selling. The output is a faster dinosaur.
Lane 2: Asteroid shelters. Defensive infrastructure sold to the same dinosaurs once they realize they are dinosaurs. This lane makes money. It also runs on a clock, because the asteroid is coming no matter how many shelters you sell, and at some point the dinosaur is extinct regardless. You are cashing checks on a finite calendar.
Lane 3: New infrastructure. Net new businesses and services that could not have been built before. Entirely new ways of solving problems that require the new capabilities to exist at all. The defining companies of the next decade live here. See the imagination economy for the shape of what gets built when execution costs collapse, and the new economy for the survivor framing of what happens to everyone in Lane 1's clientele.
Where The Wealth Goes
Ninety-nine percent of the wealth anyone makes in the next three years comes from Lane 3. That is my working estimate. It is approximate. The imbalance is real.
The reason is structural. Lanes 1 and 2 pay per dinosaur, and the dinosaurs are a finite, shrinking population with a lot of gatekeepers between you and the check. Lane 3 pays per customer served in a market that did not exist before you entered it, with no gatekeeper, at a cost basis that most incumbents cannot match. A Lane 3 business can go from zero to a hundred million in eighteen months. A Lane 1 practice can go from zero to a thousand clients over a career.
There is also the moat problem with Lane 1. Once you have helped the dinosaur standardize its tribal knowledge into a reusable harness, you have also given the dinosaur's competitors a playbook. Your own deliverable is the blueprint for the thing that replaces you. See the coming revolt against AI services companies for the version where the client realizes this and the narrative turns.
Why Operators Default To The Wrong Lane
Lanes 1 and 2 feel safer. The checks are larger and the sales cycle has a known shape. You can explain the work to your parents. Lane 3 requires you to bet on a business model that your own network cannot evaluate, because it has never existed before.
That is exactly why Lane 3 is where the money is. Everyone who could be competing with you has settled for dinosaur work.
The Practical Move
If you have capital and runway, go to Lane 3 now. If you do not, take Lane 1 or Lane 2 work on the explicit condition that every hour of dinosaur enablement is funding the Lane 3 thing you are actually building. The dinosaur work is a grant. The new infrastructure is the job.
Above all, refuse to frame dinosaur enablement as the main event. It is the funding vehicle. The main event is whatever it is you will have shipped when the dinosaur is extinct and your old client list is a graveyard.
Dinosaurs get the last of the old money. Asteroid shelter sales pay a clock. The new economy pays the operator who built the thing that could not exist before.